>As people gained experience and fed back
>ideas, the closed loop pointing system has
>evolved into an "adaptive" one which does
>pointing corrections only when needed.
In addition, empirical terms for the last bit of polar misalignment and flexure are relatively simple to build, even into a client script, and I've found these are all that is needed if the telescope is properly balanced and has no egregious problems (like Alain's gear problem). On most of my scripts at this time, I check the pointing on the data image after the fact to "make sure" things are right. Only rarely is a problem found, and I have correlated a few of those problems to an owl landing on the telescope.
>Thus, "George's Rule" (Doug George that is) holds: The
>more money you spend the more work you can get done per
>unit time.
From a fiscal planning perspective, to justify an expensive mounting on these grounds alone, you'd have to show that spending twice as much money results in more than twice as much work per unit of time than if the lesser mounting were used. This can't be done; the users of premium mountings that I know get approximately a 5% gain by skipping the checks mentioned above. It is hard to argue that this magnitude of gain is worth a ~600% increase in system cost. Especially when merely doubling the cost gets you twice the efficiency (by buying two identical systems) (which is why I say you'd need a >2 factor to justify the cost). And even more so when the big bottleneck for most users is downloading images through that crappy parallel port - speeding that up could be more cost-justifiable on an efficiency basis.
Of course, there are a number of other reasons to use a higher-end mounting (load carrying capability is #1 in my book). But if you have only $10,000 to spend on the whole system, forget it.